Pellegrini - Recess, Cognitive Performance, And School Adjustment

Recess, Cognitive Performance, and School Adjustment 1

The Role of Recess in Primary School*

Anthony D. Pellegrini Department of Educational Psychology

University of Minnesota/Twin Cities Campus and

Robyn M. Holmes Department of Psychology

Monmouth University, 16 May 05


The Role of Recess in Childrens Cognitive Performance and School Adjustment

There can be little doubt that there has been increased emphasis on accountability in both

preschool and primary school education over the past 20 years. Opinions regarding the value of

this position, however, are certainly diverse. Advocates of the accountability movement,

rightfully suggest that scarce tax dollars should be spent only on programs that work.

Politicians as different as former President Bill Clinton and President George W. Bush have

supported variants of this view.

One of the specific, though less noticed, impacts of accountability has been that

childrens opportunities for free time (in the form of recess) and corresponding opportunities to

interact with their peers has been eliminated or diminished in many school systems across this

country, Canada, and the UK (The Economist, 2001; Pellegrini, 2005). The popularity of the

movement to minimize recess in schools may be due to the fact that politicians and school

superintendents see this as a way in which to get tough on education, provide more academic

time for students, and to improve academic performance. Indeed, it may seem commonsensical

to many that reducing recess time has a positive effect on achievement. After all, more time

spent in academics should directly translate into improved performance (Brophy & Good, 1974).

This as viewed was advanced by the former superintendent of schools in Atlanta, Benjamin

Canada, even though there was no empirical or theoretical evidence for the claim (The

Economist, 2001).

On the other hand, many educators and parents, especially, recognize the centrality of

teaching skills and maximizing the efficient use of relatively scarce classroom time, but they also

see the necessity for the role of breaks between periods of intense work where children can both


relax and interact with peers, with the hope that they will return to their classrooms after their

breaks and work with renewed interest.

There can be common ground between the pro- and anti-recess positions. Many,

including the authors of this chapter, feel that there is a need for accountability. Our best theory

and empirical evidence should be used to guide practice. To do otherwise is to squander the trust

and resources of children, families, taxpayers, and educators. Indeed, far too many of the

educational policies being recommended for primary school are a form of folklore as they have

no scientific basis. Yet, what many dont realize is that recess IS educational. In this chapter

both theory and empirical data are presented to support the argument that what goes on during

the recess period is educational in the traditional sense.

Recess is defined here as an unstructured break time between periods of relatively

rigorous academic time. Further, recess can take place either indoors or outdoors and should

involve childrens free choice of activities and playmates, with minimal adult direction; the role

of adults should be to supervise childrens safety. For example, adult supervision should prevent

children from bullying their peers and using unsafe behaviors on play apparatus, such as standing

on a see-saw; only very minimally should adults structure childrens play and games. From this

view, recess is more characteristic of primary schools than preschools as the latter often have

play as part of the curriculum.

Cognitive Performance in School

An important goal of primary grades schooling, unlike many preschools and nurseries, is

teaching children skills and strategies associated with literacy, mathematics, and science -

typically measured by some form of standardized achievement test. The term cognitive

performance is used in this chapter as an umbrella term to cover those skills and strategies


associated with school-based learning. The effects of recess breaks on primary school childrens

satisfaction with school and standardized test performance, as well as more direct measures of

cognitive performance - attention to classroom tasks- are assessed. Direct measures of attention

to class work, of course, relate to more general cognitive measures, such as achievement test

performance: Students must attend to their work in order to learn it. As we know from years of

research, dating back to the 19th century (e.g., Ebinghaus, 1885; James, 1901), distributing work

over a long period of time with breaks interjected, compared to massing it into one period,

maximizes attention and more general cognitive performance across the life span.

For young children, a specific form of break from cognitively demanding tasks may be

especially effective in maximizing cognitive performance, as suggested by David Bjorklunds

cognitive immaturity hypothesis. This position holds that playful break times and

corresponding peer interaction may be especially important in maximizing performance for

primary school children to the extent that such breaks reduce the cognitive interference acquired

in earlier instruction (Bjorklund & Harnishfeger, 1987; Bjorklund & Green, 1992; Bjorklund &

Pellegrini, 2000). The playful nature of breaks is predicated on the assumption that childrens

thought processes are different from adults.

Traditional views of childrens cognitive processing (e.g., Piaget, 1983) suggest that

young childrens cognition is an imperfect version of more mature adult processes. For example,

young childrens tendencies to make unrealistic estimates of their own capabilities by over

estimating their own cognitive (Yussen & Levy, 1975) and social status (Smith & Boulton, 1990)

has been framed as a limitationsomething to be overcome. The cognitive immaturity

hypothesis suggests that these processes are not inferior variants of adult behavior but instead

specific adaptations to the niche of childhood that enable young children to effectively learn


skills and behaviors (Bjorklund & Pellegrini, 2000). For example, childrens over-estimation of

their own cognitive and social skills enables them to persevere at tasks even though, by adult

standards, they are not doing it very well. This perseverance may lead to self-perceived success

which may, in turn, lead to higher self-perceived competence and help with learning complicated

skills and strategies (Bandura, 1997).

With specific reference to the role of recess, this position holds that playful, not

structured, breaks may be especially important in maximizing performance because unstructured

breaks may reduce the cognitive interference associated with immediately preceding instruction

(e.g., Bjorklund & Harnishfeger, 1990). The immaturity of childrens nervous systems and their

lack of experiences render them unable to perform higher-level cognitive tasks with the same

efficiency as older children and adults, and directly influence their educability. From this logic, it

follows that young children are especially susceptible to the effects of cognitive interference

after sustained periods of structured work (see Dempster, 1992). Breaks during periods of

sustained cognitive work should reduce cognitive interference and maximize learning and

achievement gains (Toppino, Kasserman, & Mracek, 1991). Further, opportunities for peer

interaction during playful breaks not only enable childrens feeling of potency and competence,

but they also help children learn and develop the important social skills necessary for successful

peer relations both in and out of school.

The Role of Peer Interaction at Recess in Predicting First Grade Achievement

Playful break times as a recess break provide an interesting venue not only for children to

learn and develop social skills but they are also important for assessing young childrens

cognitive performance. Specifically, unstructured peer interaction is a cognitively and socially

very demanding context while simultaneously very motivating for children. This combination of


high cognitive demands and high motivation should lead to children exhibiting higher levels of

competence in recess activities, in comparison to traditional achievement tests, that are not as

motivating (Messick, 1983). Consider the social cognitive demands associated with sustained

peer interaction. Children must take others perspectives, communicate effectively, and follow

negotiated rules (Pellegrini, 1982). That children enjoy doing these things means that they are

motivated to do the necessary social cognitive work to accomplish such high level tasks.

Recess as a demanding and motivating context. Childrens interactions with peers during

playful interaction, relative to interactions with adults, are cognitively and socially more

demanding. A number of studies have demonstrated that, when given free choice in a play

environment, children who choose to interact with peers, relative to adults, are more

sophisticated on a number of social cognitive dimensions (e.g., Harper & Huie, 1985; Pellegrini,

1984; Wright, 1980). For example, in a university preschool sample, we found that when adults

were in childrens immediate groups childrens play and oral language production was less

sophisticated than when the groups were comprised of children alone (Pellegrini, 1984; 1985).

Childrens use of oral language with their peers, relative to language used with adults, was more

explicit (e.g., pronouns are defined) and more narrative-like (e.g., use of temporal and causal

conjunctions such as and then and because). Childrens use of this sort of literate language

reliably predicts performance on tests of early literacy (Pellegrini & Galda, 1993). Further,

childrens fantasy more frequently involved them taking a greater variety of roles, including

those, such as pretending to be the teacher, where they used language to direct their playmates --

a form of language not likely to be observed when teachers are part of the group!

Consistent with Piagetian (1983) theory, the disequilibration characteristic of peer

interaction facilitates social and cognitive development. Disequilibration is defined as conceptual


conflict between what children know and what they encounter. When faced with conceptual

conflict children try to resolve it and this resolution results in conceptual development. For

example, a boy and girl in kindergarten are playing with a doctors kit. The girl says, Ill be the

doctor and you be sick. That girls can be doctors may conflict with the boys concept of

doctors; this encounter broadens the boys concept of doctor and what females can do. When

individuals disagree, they are confronted with points of view other than their own. In order for

the interaction to continue, they must incorporate (or accommodate, to use Piagets term) the

other persons points of view. In short, engaging in sustained social interaction with peers

requires a fair amount of social (e.g., cooperation, perspective taking) and cognitive (e.g., ability

to communicate clearly) skill.

In comparison, when interacting with an adult, the adult will often take over some of this

difficult work, especially if the child is having problems (Vygotsky, 1978). Further, children do

not tend to question adults because they are typically socialized not to question or challenge

grown-ups. In short, conceptual disagreement, or disequilibration, is more common in peer

interaction than in adult-child interaction, which is typically unilateral.

Playful breaks at recess should be a particularly good place to study the ways in which

childrens social cognitive development is facilitated in playful, peer interaction contexts. As

noted above, young children are typically motivated to exhibit high levels of competence there

because they enjoy interacting with their peers. Consequently, observations in such a highly

motivating context should maximize children's exhibition of social and cognitive competence of

the sorts noted earlier (Pellegrini, 1984; Waters & Sroufe, 1983; Wright, 1980). By implication,

placing children in such a highly motivating situation is very important from an assessment

perspective. It may be that the often-described difference between children's competence as


measured in standardized testing situations and their competence measured in playful situations

is due to different levels of motivation to exhibit competence in each situation (Vygotsky, 1978).

Recess behavior in kindergarten predicting 1st grade achievement. Longitudinal analyses

also demonstrate relations between kindergarten children's social behavior at recess on the school

playground and measures of 1st grade achievement. Kindergarten recess behaviors as a predictor

of 1st grade achievement is especially important to consider because this is the transition into full

day schooling and also a period in children's development when social interaction and play may

be especially important. More specifically, Pellegrini (1992) examined the relation between

kindergarten childrens behavior at recess and their performance on a standardized achievement

test in 1st grade. The intent of this research was to show that standardized achievement and

aptitude measures alone have limited explanatory and predictive power, especially for young

children.

Zigler and Trickett (1978; see also Zigler & Bishop-Josef, this volume) made this point a

generation ago in service of a similar cause, i.e., to argue against over reliance on standardized

measures of cognition, such as IQ tests, in favor of measures of social competence, to assess the

impact of early intervention programs, like Project Head Start. Again, the basis of this argument

is that childrens optimum performance is not elicited in standardized tests. Instead, peer

interaction on the playground at recess may be a more valid assessment context because it is both

highly motivating and very demanding. Like Zigler and Trickett, the analyses presented below

should help to guide legislators and departments of education in their views of teaching and

assessing young children.

The results from a two-year longitudinal study conducted in a public school in Athens,

Georgia support the role of peer interaction at recess in childrens achievement (Pellegrini,


1992). This research was motivated in reaction to a law in the State of Georgia (where Pellegrini

was living at the time) requiring kindergarten children to pass a standardized achievement test in

order to be promoted to first grade. The school population where the research was carried out

was composed of a wide variety of children ranging socio-economically from very poor to very

affluent. Children were observed across two years on the school playground during their recess

periods. Information on their standardized, academic achievement test performance was

available. In this prospective longitudinal study, the Metropolitan Readiness Test (MRT, Nurss

& McGauvran, 1976) was used to assess the kindergarten childrens general knowledge and

achievement in early reading and math concepts. The measure of academic achievement for 1st

graders was the Georgia Criterion-Referenced Test (GCRT), also a test of general knowledge

and achievement in early literacy and numeracy concepts.

The results demonstrated that childrens recess behavior as kindergarteners was a

significant predictor of their 1st grade academic achievement, as measured by the GCRT, even

after statistically controlling their kindergarten achievement. The measures of recess behavior

used were rather gross measures - peer interaction and adult-directed behavior. The peer

interactions coded included all forms of cooperative behavior, including play, games, and

conversation. Similarly, adult-directed behavior (that is, behavior initiated by a child with an

adult) included any overtures to adults by children and positive adult-child interaction. A bar

graph, displayed in Figure 1, represents the cumulative variance accounted for in this model. The

variables entered, in order, were MRT as a statistical control for kindergarten achievement, Adult

Interaction (which was a negative predictor of 1st grade achievement), Object Play, Peer

Interaction.

/Figure 1 about here/


In step 1 the MRT control variable was entered first in the predicting 1st grade achievement and

it accounted for .34 of the variance in 1st grade achievement. Then adult-directed behavior on the

playground in kindergarten was entered and it accounted for another, unique .18 of the variance

in 1st grade achievement, but the relation was negative, not positive. Object play and then peer

interaction on the playground in kindergarten each accounted for another and unique .20 and .11,

respectively, of the variance in 1st grade achievement.

Importantly, peer interaction was positively related to achievement, while adult-directed

behavior was negatively related to achievement. That is, children who interacted more with peers

at recess tended to score higher on the GCRT than children who interacted less with their peers.

In fact, children who were involved in high levels of adult-directed behavior at recess tended to

score lower on these achievement tests than children who interacted less with adults.

These results suggest a number of interesting mechanisms that might have driven the

recess advantage. For example, children who chose to interact with adults at recess, rather than

with peers, may have lacked the social skills necessary to play with the other children. Evidence

from the social competence literature demonstrates that when young children choose to interact

with teachers, compared to peers, in play oriented contexts, teachers do most of the work in

maintaining interaction (Harper & Huie, 1985; Wright, 1980, 1980). In comparison, when

children interact with peers, they must rely on their own social competence to initiate and sustain

interaction. Relatedly, children who chose to interact with adults may have been unpopular with

their peers. Consequently, they may have been rejected by the peers and by default only had

adults with whom to interact. In either case, difficulties in peer relation often forecasts academic

difficulties (Coie & Dodge, 1998).


The important point here is that what children do at recess accounts for a statistically

significant, and unique, portion of the variation (40%) beyond what standardized tests tell us.

These findings support the arguments made by Zigler and Trickett (1978;see also Zigler &

Bishop-Josef, this volume) that standardized measures of children's cognition have limited

predictive value. Substantially more variance in 1st grade achievement is accounted for when

children's behavior in a naturalistic and motivating environment is considered. This speaks

against the current practice of using one measure (as in the case of No Child Left Behind) to

make inferences about childrens cognitive performance, especially a measure that focuses

exclusively on the academic side.

The varying role of adults in children's cognition in different contexts is consistent with

extant research. For example, in free-play situations, adults generally inhibit older preschool

children's exhibition of complex forms of play, whereas peers facilitate it (e.g., Dickinson &

Moreton, 1991; Pellegrini, 1984; Pellegrini & Galda, 1993). However, in a small-group teaching

context, such as planning an errand or a classification task, adults are much more effective tutors

than are peers (Tudge & Rogoff, 1989). In short, adult guidance is useful in some contexts, like

skill teaching, and peer interaction is important in other contexts, where social competence is at

stake.

Peer interaction as the 4th R: The importance of peer interaction for adjustment to school.

The second dimension of the argument made here relates more specially to the

importance of peer interaction for childrens more general adjustment to primary school. As

noted earlier, the observations of recess behavior that were made in the above reported study

were rather gross. These measures only told us about the relative value of peer and adult


interaction. In what follows, the role of a specific form of peer interaction, namely, childrens

games with their peers, in predicting school adjustment will be examined.

At the level of theory, the ability to interact cooperatively with peers, inhibit antisocial

behaviors, and form close relationships, such as friendships, are all important developmental

tasks for children in the primary grades (Hartup, 1996; Waters & Sroufe, 1981; Zigler &

Trickett, 1978). Developmental tasks vary with age (e.g., impulse control during the preschool

period and peer group membership during the early primary school years, see Berk, this volume)

and the successful mastering of these tasks constitutes social competence for that period and

provides the foundation for subsequent development. From this view, mastering the skills

necessary for membership in ones school peer group and feeling efficacious in this area should

provide a basis for subsequent adjustment to school.

Numerous longitudinal studies and field experiments documented the importance of

childrens peer relations in their initial adjustment to elementary school (e.g., Ladd,

Kochenderfer, & Coleman, 1996; Pellegrini, 1992; Pellegrini, Kato, Blatchford, & Baines,

2002). For example, Ladd and colleagues (Ladd, et al., 1996; Ladd, Price, & Hart, 1988)

demonstrated that successful transition from preschool to primary school is fostered when

children make the transition with a friend. Friends provide social emotional support for each

other in the new, and sometimes stressful, school environment.

One particular type of peer interaction, playing games with rules, such as tag, soccer, and

jump rope games is especially important for children making the transition to full day schooling

because these games represent a motivating and important developmental task for that age group

(Sutton-Smith, 1971). As Piaget (1965) argued, games are a modal form of interaction for

primary school-age children. Correspondingly, the design features of games are representative of


a developmental task for that period. For example, following an externally defined a priori rule is

important not only in games but also important in the larger arena of social exchange. Sutton-

Smith (1971) and others (Pellegrini et al., 2002) have documented empirically the importance of

certain forms of games (games of challenge and leadership) for children, and especially boys,

making the transition to primary school. Pellegrini and colleagues (2002) defined general

facility with games in terms of percent of time spent in games on the playground at recess as well

as peer nominations and teacher ratings of children as leaders in games. It was found that

playground games only predicted boys, but not girls, school adjustment.

This finding is consistent with the theoretical assumption that the social rules and roles

that children learn in one niche (with their peers at recess) predict competence in a related niche -

adjustment to school more generally. Both the school playground at recess and classroom

settings are similar to the extent that they encourage rule governed behavior and cooperative

interaction with peers.

The lack of relations between girls outdoor games and school adjustment may be related

to the fact that girls, generally, are less keen on playing outdoors, relative to boys (Harper &

Sanders, 1975). This relative lack of enthusiasm for the outdoors as a play venue may be due to

the fact that girls find boys loud and boisterous behavior outdoors unpleasant and thus segregate

themselves from boys into more sedentary groups (Maccoby, 1986; Pellegrini, 2004). It would

be interesting to see the degree to which girls games predict school adjustment in an all girls

school- where boys crashing bodies do not inhibit their games. Further, girls facility with

indoor games is also worth examining.

The research cited above also showed how facility with outdoor games during the first

part of the school year predicted end-of-year social competence and school adjustment even after


controlling for social competence and school adjustment, at the beginning of the year (Pellegrini

et al., 2002). Social competence in this study was defined following Sroufe and colleagues

(Sroufe, Egelund, & Carlson, 1999) at both the group (being liked by peers) and relationship

levels (reciprocal friends). That game facility predicted childrens school adjustment is a very

important finding for educational policy makers. Games probably provide opportunities to learn

and practice the skills necessary for effective social interaction with peers in an important

socialization context - early schooling. Further, being facile in one dimension of school, albeit

playing games at recess, is probably very important for childrens more general feelings of

efficacy in school. Feelings of efficacy about school may relate to subsequent achievement

motivation, for example, attendance and positive classroom participation.

While these results reinforce earlier research where childrens peer relations in school

predicted school success (e.g., Ladd & Price, 1993), they extend this earlier work in that the

majority of the students in the current study were low-income children. It is well known that

children, and again boys especially, from economically disadvantaged groups, have difficulty

adjusting to and succeeding in school (e.g., Ackerman, Brown, & Izard, 2004; Heath, 1983). We

demonstrated that their success in one part of the first grade school day (games at recess) can

predict more general school adjustment.

The Role of Recess in Cognitive Performance: Proximal Measures

In the preceding section the positive role of recess in childrens achievement tests scores

was demonstrated. Recess also has a relatively immediate impact on childrens cognitive

performance. Specifically, in this section we illustrate how childrens attention to classroom

tasks is facilitated by recess breaks. It is probably this increased attention that is responsible, in

part at least, for the positive role of recess in achievement.


Attention is a measure that is consistent with theories suggesting that breaks from

concentrated schoolwork should maximize performance. Following the notion of massed vs.

distributed practice (Ebinghaus, 1885; James, 1901), children are less attentive to classroom

tasks during longer, compared to shorter, seatwork periods (e.g., Stevenson & Lee, 1990).

Correspondingly, children are more physically active and socially interactive on the playground

after longer, compared to shorter, confinement periods (e.g., Smith & Hagan, 1980). These

breaks and the behaviors observed during breaks, in turn, have implications for children's

classroom behavior after recess (e.g., Hart, 1993), as well as more distal academic and social

cognitive development, as was demonstrated in the previous section. In the studies reported

below, attention was measured in terms of childrens looking at either their seat work or, in cases

when the teacher was reading to them, at the teacher. Additionally, in some studies childrens

fidgeting and listlessness was also coded as a measure of inattention while children did their

seatwork.

In what follow the effects of recess timing on childrens attention to classroom work was

examined. By recess timing we mean the amount of time before recess that children are forced to

be sedentary (or are deprived of social and physical play) and attend to class work. This type of

regimen typifies most primary school classrooms (Minuchin & Shapiro, 1983). The school in

which this research was conducted allowed the researchers to manipulate the times that children

went out for recess as well as what they did in their classrooms before and after recess.

Consequently, these experiments enabled researchers to make causal inferences about the effects

of recess timing on childrens attention to classroom tasks.

The children enrolled in this public elementary school were from varied social economic

and ethnic backgrounds. In all of the cases, the children in each of the grades were systematically


exposed to different schedules for recess timing. On some days they went out to recess at 10 AM

and on other days at 10.30 AM. Before and after each recess period children were read an

experimentally manipulated male-preferred (with male characters) or a female-preferred (with

female characters) book. During this time we coded their attention to the task. We also observed

and coded their recess behavior.

Our orientation in studying the role of recess timing in childrens attention suggests that

during that period immediately preceding recess, children's inattention to instructional tasks

should increase as a function of duration of the deprivation period. Anecdotal evidence from

Japan and Taiwan suggests that children's attention to class work is maximized when

instructional periods are relatively short, not long, intense, and there are frequent breaks between

these work periods (Stevenson & Lee, 1990). In these countries, as in the Peoples Republic of

China (Kessen, 1975), children are given a break every 50-minutes or so. When children return

from breaks they seemed more attentive and ready to work than before the breaks (see also

Lewis, 1995).

In the studies reported below, the effects of both indoor and outdoor recess periods on

attention are examined. Examining the effects of indoor recess on childrens attention would

provide insight into the role of a relatively sedentary break period on subsequent attention. Thus,

these results should provide further evaluation of the blowing off steam hypothesis (Evans &

Pellegrini, 1997). If childrens attention is greater after than before the indoor break, the role of

physical activity per se should be minimal. Further, and from a policy perspective, educators

sometimes use indoor recess as an alternative to outdoor breaks. This may be due to the fact that

teachers or playground supervisors are reluctant to go outdoors during inclement weather or they


may be sensitive to the possibility of law suits related to injuries on playground equipment.

Results from this experiment should provide insight into the efficacy of indoor breaks.

The first study, to our knowledge, to address directly this issue of outdoor recess activity

and post-recess attention found that 3rd grade childrens attention before recess was lower than it

was after recess, especially for boys, thus suggesting that recess facilitates attention (Pellegrini &

Davis, 1993). However, before accepting these findings at face value and assuming that recess

causes children to be more or less focused in their studies, at least one alternative explanation for

the results is possible. These results may have been due to the fact that attention was related to

the gender role stereotypicality of the tasks on which the children worked. Specifically, in the

Pellegrini and Davis (1993) study childrens class work often involved listening to a story.

Because the researchers did not systematically monitor the stories read, it may have been the

case that some of the stories read were more preferred by girls. Thus, their attention may have

been related to the task, not the effect of recess. In the next series of experiments this confound

was removed by systematically varying gender-preference of tasks before and after recess.

In Experiment 1 of a new series of experiments, the effects of outdoor recess timing on

the classroom behavior of boys and girls in grades K, 2, and 4 was examined (Pellegrini et al.,

1995). As in all experiments in this series, recess timing varied by 30-minutes. Childrens

attention (assessed by looking at the book being read to them or at the teacher doing the reading)

was assessed before and after recess on male-preferred and female-preferred books.

In Experiment 1, the pre-recess results supported the suppositions of Stevenson and Lee

(1990), who proposed that children are less attentive during long, compared to short, work

periods. That is, children were generally more attentive during the short deprivation period,

relative to the long period, and older children were more attentive than younger children were.


For example, 4th grade childrens mean attention scores were greater during the short deprivation

time, relative to the long deprivation time. It should neither be surprising that children are less

attentive as the time they spend on a task increases; nor should it be surprising that their attention

to school task increases with age. Further, children were more attentive after recess than before.

Figure 2 displays the pre-recess and post-recess inattention for grades K, 2, and 4. Note that at

each grade level the children were more inattentive before than after recess. These data are

displayed in Figure 2.


The pre-recess results also support earlier work suggesting that attention to school tasks

increases with age (Wittrock, 1986). Further, it was found that boys and girls attention to the

book read to them was influenced by the gender-role stereotypicality of the story. For example,

4th grade boys in the long timing condition were more attentive to male-preferred stories and less

attentive to female-preferred stories while the pattern was reversed for the girls. This finding is

consistent with the extant literature on gender preference for stories (Monson & Sebesta, 1991).

Results from this experiment should be interpreted cautiously primarily because of the

small sample size (20 children/grade and 10 children/sex within each grade) and because there

was only one classroom at each grade level. Replication is clearly needed to assure that the

results are not aberrational (Lykken, 1968), especially when the results have implications for

school policy. Replication is also needed to clarify the effect of condition on physical activity at

recess; the results from this experiment were not consistent with the one other experiment

involving primary school children. With these needs in mind two more experiments were

conducted (Pellegrini et al, 1995).


In Experiment 2, the same outdoor recess timing and attention procedures were used as in

Experiment 1. In Experiment 2, 2nd and 4th graders (one classroom for each grade) were studied

in the same school as in Experiment 1. The results from Experiment 2, similar to those from

Experiment 1, revealed that children's task attention is affected by recess timing and that timing

interacted with dimensions of the task as well as children's age and gender. Children generally,

but especially 2nd graders, were more attentive after recess, relative to before recess. In addition,

a significant difference between pre-recess and post-recess attention was observed for 2nd grade

only, not for 4th grade. That is, only second graders, not 4th graders were more attentive after

recess that before. In Figure 3, the pre- and post-recess inattention scores are displayed. The

values 1 and 2 indicate 2nd grade pre-recess and post-recess inattention, respectively. Similarly, 3

and 4 represent 4th grade pre-recess and post-recess inattention, respectively. In both cases note

that inattention is greater after recess than before.


In Experiment 3, students in two 4th grade classrooms were studied and the same recess

timing paradigm employed in the previous experiments was used. The recess period was indoors,

however. The same experiment was conducted with two separate intact classrooms. Such a

design was chosen because of the relatively small samples involved in each classroom. This

procedure minimizes the probability of obtaining aberrant results if similar results are obtained in

both samples - thus replicating each other. The results from this experiment are similar to those

from other experiments: Attention was greater after the recess period than it was before the break

in classroom 2. The result from the indoor recess study was similar to the outdoors results-

Children are generally more attentive to classroom work after recess than before. Whether recess

is indoors or outdoors does not seem to matter. In Figure 3, these results are displayed with pre-


recess (Numbers 1 and 3) and post-recess scores (Number 2 and 4) of inattention for the two

classrooms. Again, note that in both cases, inattention is lower after recess than before, as

displayed in Figure 4.


The resulting message from all of this research is clear. Breaks between periods of

intense work maximize childrens attention to their class work. More speculatively, it is probably

this increased attention that is partially responsible for the positive relations between recess and

performance on achievement tests.

What we really still need to know, however, is the degree to which different types of

recess regimens relate to attention and achievement. A first step in this direction was already

presented, as we found that indoor recess periods, like outdoor periods, seemed to be effective

facilitators of childrens attention to class work. This result is consistent with mass vs. distributed

practice as the mechanism responsible for the effects of recess on attention. From this view, the

nature of the break is less important than having a break per se. To more thoroughly examine this

explanation, researchers should examine different types of breaks after periods of intense

cognitive work. For example, does watching a short video or listening to music for a short period

facilitate attention? Further, do these effects vary with the age and gender of the child?

Bjorklunds cognitive immaturity hypothesis provides some guidance here (Bjorklund &

Pellegrini, 2000). The theory suggests that breaks for preschool and young primary school

children should be playful and unstructured. Providing time for children to interact with peers

or materials on their own terms, that is, with minimal adult direction, should maximize attention

to subsequent tasks. The venue, indoors or outdoors, is not important. With older children,


merely providing breaks between periods of intense work might suffice to maximize attention.

The research on massed vs. distributed practice with adults supports this view.

Another crucial aspect of the recess period that needs further research is the duration of

the recess period. Should it be 10 minutes, 20 minutes, 30 minutes, etc? We simply do not know.

Answers to this question have obvious implications for school policy and scheduling. We

explored the issue of recess duration with a sample a preschool children (age 4- 5 years)

(Holmes, Pellegrini & Schmidt, unpublished data). Procedurally, children had circle time and

were read a story before recess. Then they went outdoors for recess periods of 10, 20, or 30

minutes. When they returned to their classrooms they again sat in a circle and listened to a story

read to them by their teachers. Attention was recorded and coded (using scan sampling and

instantaneous recording procedures) for whether the child was gazing in the direction of the book

or the teacher reading the book.

Consistent with earlier work with older, primary school children (Pellegrini & Davis,

1993, Pellegrini et al., 1995), attention was greater following, relative to before, the recess period

and girls were more attentive than boys to classroom tasks in all conditions. Thus it seems

reasonable to conclude that outdoor recess breaks help children attend to classroom tasks.

Regarding the differing durations of recess periods, we found that attention to classroom

tasks was greatest following the 20- and 10-minute outdoor play periods whereas the 30-minute

period resulted in higher rates of inattention. These findings are consistent with anecdotal

evidence from the UK that stated that children become bored with recess after too long a period,

and thus longer periods may become counter-productive (Blatchford, 1998).. This work needs to

be replicated as it represents only one short term study with a relatively small sample. The

duration of the recess periods is clearly important as it is one of the persistent questions posed to


us by parents and teachers. How long should the recess be? Should its duration vary with

childrens age? Clearly at this point we only have hints, and are not sure.

While childrens attention to classrooms instruction is obviously very important we

should not lose sight of the importance of peer interactions for childrens successful development

as both students and members of society. The importance of developing peer relations is

especially important during the preschool and early primary school years (3- 6 years of age)

(Waters & Sroufe, 1983). Thus, providing recess periods that encourage opportunities for

relatively unfettered peer interaction as well as exploration of different materials would

maximize both the social and cognitive benefits of recess.

Policy Implications

Data presented in this chapter provide empirical support for the positive role of recess in

the school curriculum. Providing breaks over the course of the instructional day facilitates

primary school childrens attention to classroom tasks. That these results were obtained using

well-controlled field experiments and replicated a number of times by different groups of

researchers should provide confidence in the findings. In terms of specific class performance, the

results presented here support the anecdotal evidence from Taiwanese and Japanese schools

suggesting that in order to maintain high levels of attention, children need frequent breaks in the

course of the day (Stevenson & Lee, 1990). Recall, Stevenson and Lees study compared the

achievement of Asian and American school-age children and found Americans lagging far

behind. They found that Asian schools had longer school days and school years but they also

provided children with breaks every 50 minutes across the entire school day.

This evidence should inform policy and be used to guide those policy makers and

politicians seeking to diminish or eliminate recess from the school curriculum. First, and like


Asian schools, American childrens school days and school years should be lengthened. Such a

policy would bring the total number of hours American children attend school in a year closer to

other countries in the world. This increase should have the corresponding benefit of increasing

achievement. Longer school days and years- with frequent breaks- should accomplish this.

Additionally, a longer school day and year would ensure that children have a safe place while

their parents are at work. Children would be in school, learning and interacting with their peers,

rather at home, unsupervised, or in expensive after-school care.

While recess periods alone seem to be important, it is also important to let children

interact on their own terms with peers. Providing children with a physical education class as a

substitute for recess, does not serve the same purpose (Bjorklund & Pellegrini, 2000; Council on

Physical Education for Children, 2001). They need to PLAY! This conclusion is supported on a

number of fronts. First, and at the level of theory, both the cognitive immaturity hypothesis and

massed vs. distributed practice theory suggest that after periods of intense children need breaks

from instruction. A physical education class is another form of instruction and thus would not

provide the sort of break need to maximize instruction. Further, physical education classes

typically do not provide the variety of rich opportunities for peer interaction that recess does. The

research presented above pointed to the importance of peer interaction for both social and

cognitive outcomes.

Second, the importance of physical education for classroom attention is typically

predicated on the idea that children need to blow off steam after periods of work and the

vigorous activity associated with physical education should serve this purpose. As noted earlier,

the idea of blowing off steam is rooted in Surplus Energy Theory- an invalid 19th century

theory (Evans & Pellegrini, 1997).


Third, and related to this last point, earlier research on the role of recess in childrens

attention found no empirical relation between the levels of childrens vigorous activity on the

playground at recess and subsequent attention to classroom tasks (Pellegrini & Huberty, 1993).

Consequently, these results showed be used against any effort to use physical education as a

substitute for recess.

Another implication of this work is that gender-preference of classroom tasks is

important in children's attention to those tasks. Specifically, research on childrens attention

consistently shows that boys are less attentive to classroom tasks than girls (Pellegrini et al.,

1995). Further, recess breaks are especially effective in maximizing boys attention, relative to

girls. These findings have implications for school-age boys being diagnosed with attention

deficit hyperactivity disorder (ADHD) (Pellegrini & Horvat, 1995). We know, for example, that

providing opportunities more breaks minimizing fidgeting and maximizes attention- two

dimensions of ADHD. Teachers and parents should be aware of the fact the classroom

organization may be responsible for their sons attention and fidgeting and that breaks may be a

better remedy than Ritalin.

Lastly, the finding that childrens competence develops in the context of interacting with

peers is especially important as children are rapidly losing these opportunities. There are signs in

both the US and the UK that children of primary school age have less time available to out of

school for interacting freely with peers and thus developing social skills and competence

(Blatchford, 1998). For example, after school many American children enter empty homes,

waiting for their parent(s) to return from work (Steinberg, 1986). There is also a trend in both the

US (Pellegrini, 2005) and in the UK (Blatchford, 1998; Blatchford & Sumpner, 1998) for recess

time to be limited or eliminated from the primary school day. Recess may be one of the few


times during the day when children have the opportunity to interact with peers and develop

social skills free from adult intervention.


We acknowledge Mark Van Ryzin for help with constructing the figures in this

chapter. The ideas expressed in this chapter also benefited from numerous

discussions on the topic across number of years with Peter Blatchford, Dave

Bjorklund, and Peter Smith. We also acknowledge support provided by the

Spencer Foundation for a portion of the research reported in the chapter.


Figure 1

Predicting 1st Grade Achievement

00. 10. 20. 30. 4

M R T A d u l t s O b j e c t s P e e r sKindergarten Variables Entered in

O rder: M R T , A du lt I nterac tio n, O bj e c t P la y , P eer I nterac tio n

Varia

nce Ac

coun

ted for in

1st G

rade

Ach

ieve

men

t

S e r i e s 1S e r i e s 2


Figure 2

K 2 4

0

5

10

15

Grade

Inattention Pre- and Post-Recess

P o s t - R e c e s sP r e - R e c e s s


Figure 3

24

01234567

Grade




Figure 4

12

01234567

C l a s s ro o m




References

Ackerman, B., Brown, E., & C. Izard, (2004). The relations between contextual risk, earned

income, and the school adjustment of children from economically disadvantaged families.

Developmental Psychology, 40, 204-216.

Bandura, A. (1997). Self-efficacy: The exercise of control. New York: W H Freeman.

Bjorklund, D. F. & Green, B. L. (1992). The adaptive nature of

cognitive immaturity. American Psychologist, 47, 46-54.

Bjorklund, D. F., & Harnishfeger, K.K. (1987). Developmental differences in the mental effort

requirements for the use of an organizational strategy in free recall. Journal of

Experimental Child Psychology, 44, 109-125.

Bjorklund, D. F., & Pellegrini, A. D. (2000). Child development and evolutionary psychology.

Child Development, 71, 1687-1708.

Blatchford, P. (1998). Social life in school. London: Falmer.

Blatchford, P., & Sumpner, C. (1998). What do we know about break time? Results from a

national survey of breaktime and lunchtime in primary and secondary schools. British

Educational Research Journal, 24, 79-94.

Brophy, J., & Good, T. (1974). Teacher-student relationships: Causes and consequences.

New York: Holt, Rinehart, and Winston.

Coie, J. D., & Dodge, K. A. (1998). Aggression and antisocial behavior. In N. Eisenberg (Ed.).

Manual of child psychology, Vol. 3, Social, emotional, and personality development (pp.

779-862). New York: Wiley.


Council on Physical Education for Children (2001). Recess in elementary schools: A position

paper from the National Association for Sport and Physical Education. Retrieved

September 28, 2000 from http://eric.ed.uiuc.edu/naecs/position/recessplay/html.

Dempster, F. N. (1992). The rise and fall of the inhibitory mechanism: Toward a unified theory

of cognitive development and aging. Development Review, 12, 45-75.

Dickinson, D., & Moreton, J. (1991, April). Predicting specific kindergarten literacy skills from

three-year-olds preschool experience. Paper presented at the Biennial Meeting of the

Society for Research in Child Development, Seattle, WA.

Ebinghaus, H. (1885/1964). Memory. New York: Teachers College Press.

Economist, The. (June 16, 2001). No time for play, p. 35.

Egertson, H. (2003). The shifting kindergarten curriculum. ERIC Clearinghouse on Elementary

and Early Childhood Education, 1-4.

Elkind, D. (1993). Images of the young child. Washington, DC: National Association for the

Education of Young Children.

Evans, J., & Pellegrini, A. D. (1997). Surplus Energy theory: An endearing but inadequate

justification for break time. Educational Review, 49, 229-236.

Harper, L.,& Huie, K. (1985). The effects of prior group experience, age, and familiarity on the

quality and organizational of preschoolers' social relations. Child Development, 56, 704-

717.

Harper, L., Sanders, K. (1975). Preschool children's use of space: Sex differences in outdoor

play. Developmental Psychology, 11, 119.

Hart, C. (1993). Children on playgrounds: Applying current knowledge to future practice and


inquiry. In C. Hart (Ed.). Children on playgrounds: Research perspectives and

applications (pp. 418-432). Albany: SUNY Press.

Hartup, W. W. (1996). The company they keep: Friendships and their developmental

significance. Child Development, 67, 1-13.

Heath, S. (1983). Ways with words. New York: Cambridge University Press.

Hirsch-Pasek, K. (1991). Pressure of challenge in preschool: How Academic environments

affect children. New Directions for Child Development, No. 53, 39-46.

James, W. (1901). Talks to teachers on psychology: And to students on some of lifes ideals. New

York: Holt.

Kessen, W. (Ed.) (1975). Childhood in China. New Haven: Yale University Press.

Ladd, G. W., Kochenderfer, B. J., & Coleman, C. C. (1996). Friendship quality as a predictor of

young childrens early school adjustment, Child Development, 67, 1103-1118.

Ladd, G., & Price, J. (1993). Play styles of peer accepted and peer rejected children on the

playground. In C. Hart (Ed.), Children on playgrounds: Research perspectives and

applications ( pp. 130-161). Albany, NY: SUNY Press.

Ladd, G., Price, J., & Hart, L. (1988). Predicting preschoolers' peer status from their

playground behavior. Child Development, 59, 986-992.

Lewis, C. (1995). Educating hearts and minds: Reflections on Japanese preschool and

elementary education. New York: Cambridge University Press.

Lykken, D. (1968). Statistical significance in psychological research. Psychological Bulletin,

70, 151-159.

Maccoby, E. E. (1998). The two sexes: Growing up apart, coming together. Cambridge, MA:

Harvard University Press.


Messick, S. (1983). Assessment of children. In W. Kessen (Ed.) Handbook of Child

Psychology, Vol. 1, History, Theory, and Methods, (pp. 477-526). New York: Wiley.

Minuchin, P., & Shapiro, E. (1983). The school as a context for social development. In E. M.

Hetherington (Ed.) Manual of Child Psychology, Vol. 4. (pp. 197-274). New York:

Wiley.

Monson, D., & Sebesta, S. (1991). Reading preferences. In: J. Flood, J. Jensen, D. Lapp, & J.

Squire (Eds.). Handbook of research on teaching the English language arts (pp. 664-

673). New York: Macmillan.

Nurss, J., & McGauvran, M. (1976). Metropolitan Readiness Tests: Levels I and II Test and

Teacher's Manual. New York: Harcourt, Brace, Jovanovich.

Pellegrini, A. (1982). Explorations in preschooler's construction of cohesive test in two play

contexts. Discourse Processes, 5,101-108.

Pellegrini, A.D. (1984). The social cognitive ecology of preschool classrooms. International

Journal of Behavioral Development, 7, 321-332.

Pellegrini, A.D. (1985). Relations between preschool children's play and literate behavior. In L.

Galda & A. Pellegrini, (Eds.), Play, language, and story: The development of children's

literate behavior. Norwood, N.J.: Ablex.

Pellegrini, A.D. (1992). Kindergarten children's social cognitive status as a predictor of first

grade success. Early Childhood Research Quarterly, 7, 565-577.

Pellegrini, A. D. (2005). Recess: Its role in development and education . Mahwah, NJ: Erlbaum.

Pellegrini, A.D., & Davis, P. (1993). Relations between childrens playground and classroom

behaviour. British Journal of Educational Psychology, 63, 88-95.


Pellegrini, A.D., & Galda, L. (1993). Ten years after: A re-examination of the relations between

symbolic play and literacy. Reading Research Quarterly, 28, 162-175.

Pellegrini, A.D., Huberty, P.D., & Jones, I. (1995). The effects of recess timing on childrens

playground and classroom behaviors. American Educational Research Journal, 32, 845-

864.

Pellegrini, A.D., Kato, K., Blatchford, P., & Baines, E. (2002). A short-term longitudinal study

of childrens playground games across the first year of school: Implications for social

competence and adjustment to school. American Educational Research Journal, 39, 991-

1015.

Piaget, J. (1965). The moral judgment of the child. New York: Free Press.

Piaget, J. (1983). Piaget's theory. In W. Kessen (Ed.) Handbook of child psychology: History,

theory, and methods (pp. 103-128). New York: Wiley.

Smith, P. K., & Boulton, M. (1990). Rough-and-tumble play, aggression, and dominance:

Perception and behavior in children's encounters. Human Development, 33, 271-282.

Sroufe, L.A., Egelund, B., & Carlson, E.A. (1999). One social world: The integrated

development of parent-child and peer relationships. In W.A. Collins & B. Laursen (Eds.),

Relationships as developmental contexts. The Minnesota symposia on child psychology,

Vol. 30 (pp. 241-261). Mahwah, NJ: Erlbaum.

Steinberg, L. (1986). Latchkey children and susceptibility to peer pressure. Developmental

Psychology, 22, 433-439.


Stevenson, H. W. & Lee, S.Y. (1990). Concepts of achievement. Monographs for the Society for

Research in Child Development Serial No. 221). 55(1-2).

Sutton-Smith, B. (1971). A syntax for play and games. In R. Herron & B. Sutton-Smith (Eds.),

Child's play (pp. 298-310). New York: Wiley.

Toppino, T. C., Kasserman, J. E., & Mracek, W. A. (1991). The effect of spacing repetitions on

the recognition memory of young children and adults. Journal of Experimental Child

Psychology, 51, 123-138.

Tudge, J. & Rogoff, B. (1989). Peer influences on cognitive development: Piagetian and

Vygotskian perspectives. In M. Bornstein & J Bruner (Eds.). Interaction in human

development (pp 17-40). Hillsdale, NJ: Erlbaum.

Vygotsky, L. (1978). Mind in society. Cambridge, MA: Harvard University Press.

Waters, J., & Sroufe, L.A. (1983). Social competence as a developmental construct.

Developmental Review, 3, 79-97.

Waters, E., & Sroufe, L. A. (1983). Social competence as developmental construct.

Developmental Review, 3, 79-97.

Wittrock, M. (1986). Students' thought processes. In M. Wittrock (Ed.), Handbook of research

on teaching (pp. 297-314). New York: Macmillan.

Wright, M. (1980). Measuring the social competence of preschool children. Canadian Journal

of Behavioral Science, 12, 17-32


Yussen, S. R., & Levy, V. M. Jr. (1975). Developmental changes in predicting ones own span of

short-term memory. Journal of Experimental Child Psychology, 19, 502-508.

Zigler, E., & Trickett, P. (1978). I.Q., social competence, and evaluation of early childhood

intervention programs. American Psychologist, 33, 789-798.

Documents

Pellegrini - Recess, Cognitive Performance, And School Adjustment